(Industrial) Internet of Things. Edge and Cloud Analytics
If you are an operator of a large pool of capital intensive assets (such as an energy producer or operator of multiple facilities), it is paramount that you have the ability to efficiently collect and analyse data in real time, devise a short-time response or use that analysis as an input to a longer term asset strategy process.
The ultimate objective is to increase efficiency and operational reliability but also to work with your ecosystem collaborators (vendors etc) in order to increase supply chain performance and minimise downtime. For instance, a typical LNG plant facility loses around $150M in unplanned downtime, an amount that certainly could be put to better use. The usual benchmark of production losses due to unscheduled events is around 5.2%.
This article is written for the decision maker (rather than for the technical gurus) by taking a wider enterprise view of how you deal with the IIOT while attempting to shed some light on the overall IIOT strategy.
the IIOT should be viewed as a
new way of thinking more than anything else.
First it leads me to define what is the meaning of “value” for my operating assets then formulate a plan (or strategy) on how to deal with the vast amount of data generated every day in order to maximise “value”. This can be done individually at the asset level (say a plant or refinery) or at the enterprise level.
VALUE means different things to different people.
The leadership of the company perceive as valuable the ability to take informed decisions along the business continuum and so, the IIOT solution is viewed as a decision support system that provides visibility across the entire enterprise.
The individual BU and functional management usually perceive value in the ability to quickly extract, synthesize data for the purpose of a specific outcome (productivity, efficiency, reliability, collaboration, safety etc).
But no matter how value is perceived, the success of an IIOT strategy is dependent on the ability of the business to create, develop and sustain new collaborative work processes across all functional lines, BU and with external ecosystem of service providers.
When it comes to analysing and synthesizing data,
one of the questions that arises is where to carry out the analytics - ultimately where do you innovate?
At the edge (closer to the asset, in almost real time) or in the cloud and why?
Or perhaps using a combination of both?
The difference between the edge and cloud analytics is overall similar to a game of football. In a soccer game the 1-3 players closest to the ball can calculate in real time the best course of action (similar to edge analytics) as opposed to sending data to the coach who then advises from the bench what needs to be done. But, nevertheless, in a post-match analysis, the coach has the ability to analyse the game further, obtain valuable input, look at a wider variety of moves, tactics and matches and advise
a.) what would have been the right course of action and
b.) suggest and implement performance improvements.
As a decision maker, some of the questions you may wish to ask are:
1.) How important is the processing or analysis time (network latency)? Do you need a real time (or near real time) response, or do you use the process for instance for a wider predictive maintenance strategy across multiple assets.
2.) What is the importance, significance and amount of data transferred (from edge/gateway into the cloud); how would you rate your ability to access, improve and standardise data?
3.) Is your business better served by sensory analytics or do you require big data deep analytics?
4.) Does your technology partner understand industrial automation and does he have a record for realising/driving value for similar clients.
5.) Do you possess a set of collaborative processes that work well across business units and function so that the organization is evolutionary fit?
The sharing and collaboration aspect of your IIOT philosophy - the social feature if you wish - is an important aspect of your strategy as it describes how employees interact with the organization to create social capital and causal ambiguity. This has enormous implications on the competitive advantage front.
In most of the cases, a combination of pre-processing data at the edge with associated deep analytics and open collaboration in the cloud (APC, CBM, enterprise data historian, planning) is the right option. If you are an oil and gas producer with a large number of discrete assets, the long-term predictive maintenance could be done in the cloud as it falls under your enterprise maintenance strategy umbrella.
The decision makers understand that asset data evolve in time and that the complexity and manipulation of such data requires integrated platforms that can seamlessly (and holistically) deal with the edge, cloud and interfaces between them.
But most importantly they realise that a sound IIOT strategy can re-orient a company’s source of competitive advantage away from physical assets toward information and intangibles (such as ecosystem collaboration with vendors and service providers, etc).
Starting with the end in mind and
choosing the right technology partner are always wise choices.